densation from the radiators and has its own condensation dripped at frequent intervals. The return mains for the "one-pipe relief" system are usually run along the walls near the floor and as nearly under the risers as practical. The return should be a wet return, but if a dry return is absolutely necessary the drips should all be water-sealed before connecting thereto. For jobs where the quantity of radiation goes up to, say, over 600 sq. ft. per riser, it is better to use the overhead or "Mills" system. This system and the "one-pipe relief" system more nearly approach a two-pipe job, having both steam main and a return main, but can be made in the one-pipe variety by using one radiator connection, with a single valve. A diagram of the well-known Mills system is shown in Fig. 3 and it will be noted that all the steam is carried up one main riser to overhead mains run around the top of the building and pitched down from the main riser to the various drops. The drops have tees at the required points of radiator supply and the condensation from the main at the top, from the riser itself and from the radiators connected thereto, falls down through the, drops-flowing with the steam all the way-and is finally drained off through drips connected to the bottom of the drops and going to the main return-preferably a wet return-through which they find their way back to the boiler. The Mills system, as can readily be seen, establishes more ideal conditions and should be used for higher building jobs, since the only portion of the piping in which the condensation is flowing against the return is in the radiator branches between the drops and the radiators. On the other hand, the Mills is a more expensive system to install than either the "one-pipe circuit" or the "onepipe relief," but in spite of this should be used for all larger work where the other two systems would be liable not to give satisfaction. TYPICAL DETAILS OF PIPING. Having the general idea of the three schemes in mind, a few typical details will go far to assist in making a perfectly-operating plant. A system which is theoretically correct, but wrong in small details, never gives satisfactory service any more than a system rightly installed in the details, but theoretically. wrong in the general idea. In all one-pipe work the steam connection to the radiator is practically alike; the connections must be made at the bottom and must be controlled (as shown in Figs. 4 and 5) by a gate valve or angle globe valve looking down, both of these points being determined by the necessity of accommodating the flow of the condensation out of the radiator. The automatic air valve should be placed on the opposite end of the radiator from the steam valve and about one-third to twothirds of the height of the radiators. above, the bottom. The runouts, of course, must be arranged to provide for expansion in the riser. mains to the risers or radiators in a "circuit" system should be inclined down from the radiator or riser to the main and should enter the main at 45°, as shown in the detail drawing, Fig. 7, thus allowing the return condensation to flow For the "circuit" system (see Fig. 1) where the steam main drops down under the water line, an automatic air valve should be placed to relieve the air caught between the advancing steam and the water seal, when starting up the system. Otherwise, the air so pocketed will interfere with the return of the condensation and cause all sorts of hammering and trouble. Best results are obtained when the air valve is installed as shown in the detail, Fig. 6. In the "relief" and mills systems, no such air valves are necessary but are sometimes used at the end of the large steam mains merely as an aid to help relieve the air more quickly when starting up. The steam runouts from the steam SIZING OF PIPES. Having the general scheme of piping and the proper details for connections the next point is to determine on the proper size of pipe. It is easy to see that a riser feeding steam up against the down-coming condensation, in order to avoid excessive velocity, must be larger than a riser on the Mills system, where no such opposing movement exists. Also, that a main carrying condensation and SRiser Runout Steam Mainz FIG. 7.-CORRECT METHOD OF MAKING RUNOUT FROM STEAM MAINS TO RISERS OR RADIATORS. steam-such as the steam main in the "circuit" system-must be made larger than one supplying steam only. As is well known, steam at atmospheric pressure, when condensed into water, occupies only a very small proportion of its original volume; in fact, it is roughly correct to say that a cubic foot of steam gives a cubic inch of water when condensed, or its liquid volume is about 1/1728 of its volume as a gas. If the steam is assumed as flowing at a rate, say, of 2,000 ft. per minute and the return at a rate, say, of 100 ft. per minute, and 2,000 cu. ft. of steam per minute is being handled, then the area of pipe for the steam will be 2000/2000, or 1 sq. ft. The steam when condensed M2Riser Runout Steam Main FIG. 8.-WRONG METHOD OF MAKING RUNOUT CONNECTION. Steam Main Riser Runout 11 Is Drip to Return FIG. 9.-RUNOUT FOR A "RELIEF" SYSTEM will occupy roughly 2000/1728: 1.15 cu. ft. 1.15 cu. ft. at 100 ft. velocity per minute equals 1.15/100 0.01 sq. ft. of pipe area. Therefore, the actual increase required in the size of steam main is next to nothing, but in the "circuit" system it is customary to run the main either full size to the end drop or nearly full size, reducing at the most not over two pipe Steam Main Runout Drop to Radiators Lowest Rod Branch. Drip to Return Mainz FIG. 10.-RUNOUT FROM STEAM MAIN IN "MILLS" SYSTEM. sizes. With the "relief" system and also with the Mills, owing to the condensation not being carried in the steam main, no consideration need be given at this point. In all cases where the flow of condensation is against that of the steam-such as in radiator branches and riser runoutsthe pipe is made one size larger than the steam supply which would be used for a common two-pipe system. Therefore, in order to size a one-pipe job it is necessary to base the sizes on those which would be used for a twopipe system; the determining factor in all steam pipe sizing is the allowable drop in steam pressure in which the job permits. For example, a system in which. the drop in pressure can be made 5 lbs. will utilize smaller pipe sizes and will be more economical to install than the same system designed for a 2-lb. drop owing to the fact that the lower drop requires larger size pipe to reduce the pipe friction to the extent required. ft. As a general thing, it is customary to allow less pressure drop and to keep 18 or 24 in. between the normal waterline and the high water-line established when the system is in operation. In other words, the lowest point of the steam and the radiation is kept at least 24 to 30 in. above the water-line of the boiler. In Figs. 1, 2 and 3 the normal water-line (N.W.L.) and the high water-line (H.W.L.) is indicated at the point on the return connections farthest from the boiler. The high water-line should always be calculated to keep at least 6 in. below the steam main or lowest level of radiation, so that, with 30 in. difference, the water rise could not be over 24 in., with 24 in. difference the rise could not be over 18 in., etc. If 1 lb. 2.3-ft. rise, then 16 oz. 2.3 X 12, or 28 in. and 1.7-in. rise. This results in a Fig. 11 illustrates a diagrammatic elevation of a one-pipe system, assuming a 1-lb. drop in the steam line between the boiler and the end of the line in times of maximum load. Thus, with 5 lbs. at the boiler and 1 lb. loss in the line, the steam pressure at the far end of the line would be 4 lbs. The water in the wet return then has a pressure of 5 lbs. per square inch pressing down on its surface in the boiler and a pressure of only 4 lbs. per square inch pressing down on its surface in the vertical leg connecting table for allowable rises as follows: with the steam main at the end of the run. This unbalanced pressure must equalize itself and does so by the water rising in the vertical leg until the vertical column of water in the leg counterbalances by its weight the difference between the two pressures; i.e., the pressure loss in the steam pipe. In Fig. 11 this loss is assumed arbitrarily as 1 lb. and since the weight of a vertical column of water is 0.43 lbs. per square inch for every foot of height, the rise of the water, under the assumed conditions, will be 1 lb. 0.43, or 2.33 • 5lbs. 1 oz. TABLE I. Allowable Rise of 3.5 7 14 28 42 70 84 Allowable Loss in Steam Mains, Pounds. 1/8 1/4 1/2 1 12 2 212 3 FIG. 11.-ILLUSTRATION OF PRESSURE LOSS IN ONE-PIPE SYSTEM. with 1 lb. drop, would require a 12-in. pipe, but, with a -lb. drop, would require a 2-in. pipe. All indirect radiators and other surfaces not equivalent to a radiation of 250 B.T.U. per square foot per hour should be reduced to such equivalent direct radiation (E.D.R.) and sized on the basis of E.D.R. Thus, an indirect radiator heating 1,000 cu, ft. of air per minute from 0° to 70° F., would give off B.T.U. as follows: 1,000 X 60 X (70° 0°) 55 B.T.U., roughly. = - = 76,000 76,000/250 304 E.D.R., and at 1⁄2 lb. drop, would require 12-in. pipe. On all horizontal branches where the TABLE II. Drop in Pressure-Pounds per Square Inch. |